Skip to content

Topic

Label Scarcity In Machine Learning

The supervised-learning problem of obtaining ground-truth outcomes for past decisions. When feedback arrives for only some decisions, retraining draws on a biased sample of the model's own behaviour.

Current clusters